Author: Alexander Barkov
Email: [EMAIL PROTECTED]
Message:
Actually it's possible to implement this without significant
efforts. Take a look into unicode.c, it has tolower unicode
mapping. In your case tolower array should also remove
accents. However, this is not so requested thing, we'll
hardly add it in near future. Feel free to hack yourself
and send us patches.

> Yes, new versions of MS SQL-Server can specify character collation and case/accent 
>sensitivity on a per column basis.
> 
> But I don't think I can do a similar thing with Oracle 8.1.
> 
> What do you think would be the cleaner way to solve that problem? And are others 
>concerned by that problem and would like to have it in mnoGoSearch?
> 
> (1) The simpler but less versatile: write a SQL script that remove accents from the 
>DICTx tables. This script is started after the indexer cron job.
> Then change the search screen to remove accents from user queries.
> 
> (2) Change indexer code to add a "NoAccent" runtime flag to remove the 
>accents before inserting the words in database. This would need to allow the use of a 
>collation map...
> 
> (3) Use a trick with LocalCharset=US-ASCII in indexer.conf to force 7 bits 
>characters?
> 
> (4) Set ispell use with a modified French dictionary?
> 
> (5) Other better way, usable in other languages...
> 
> At the present time, I use solution (1) because we already have modified the search 
>forms.
> 
> Pierre M�tras

Reply: <http://www.mnogosearch.org/board/message.php?id=3825>

___________________________________________
If you want to unsubscribe send "unsubscribe general"
to [EMAIL PROTECTED]

Reply via email to